Isolation of chromosome 21-specific yeast artificial chromosomes from a total human genome library.

A new approach for the isolation of chromosome-specific subsets from a human genomic yeast artificial chromosome (YAC) library is described. It is based on the hybridization with an Alu polymerase chain reaction (PCR) probe. We screened a 1.5 genome equivalent YAC library of megabase insert size with Alu PCR products amplified from hybrid cell lines containing human chromosome 21, and identified a subset of 63 clones representative of this chromosome. The majority of clones were assigned to chromosome 21 by the presence of specific STSs and in situ hybridization. Twenty-nine of 36 STSs that we tested were detected in the subset, and a contig spanning 20 centimorgans in the genetic map and containing 8 STSs in 4 YACs was identified. The proposed approach can greatly speed efforts to construct physical maps of the human genome.

The toxic fraction of gliadin digests in coeliac disease. Isolation by chromatography on Biogel P-10.

Improvement in the fractionation of gliadin digests and in the isolation of toxic fractions was achieved using chromatography on Biogel P-10. Fraction V, one of the 11 fractions eluted from a peptic-tryptic digest of crude gliadin extracted from Cappelle wheat, significantly affected coeliac jejunal mucosa in organ culture. BV, gamma V, omega V, the corresponding fractions V from beta-, gamma- and omega-gliadins, displayed similar toxic effects. Fraction VI containing peptides with a lower molecular mass did not show any significant cytotoxic activity and, moreover, inhibited the toxicity of fraction V. Analysis of the toxic fractions V showed that they contained peptides of 7-8,000 molecular mass, rich in proline and glutamine and poor in aromatic amino acids and carbohydrates. Among the various fractions, V, beta V from beta-gliadin appeared the less heterogeneous.

Separation of pure toxic peptides from a beta-gliadin subfraction using high-performance liquid chromatography.

The beta v subfraction was isolated from peptic-tryptic digests of beta-gliadin by chromatography on Biogel P-10 and applied to a Lichrosorb RP-18 or a mu-Bondapak C-18 column. Fractionation was achieved using reverse-phase high-performance liquid chromatography with a linear gradient of acetonitrile in ammonium acetate. A better resolution was obtained with the mu-Bondapak column. The first-eluted peptides a, b and c1 appeared to be well purified and apparently uncontaminated. Analysis of peptides a and b showed that they contained 40 to 42% glutamine/glutamic acid, 20 to 23% proline, 14 to 16% valine and 8 to 10% leucine. They had valine as the N-terminal amino acid and their molecular mass was estimated as 5500 using sodium dodecylsulfate electrophoresis after dansylation. Peptide c1 differed from peptides a and b in containing less valine and leucine and additional amino acids such as threonine, phenylalanine and tyrosine. In addition, it had a lower molecular mass (approximately 5000) and serine as the N-terminal amino acid. Peptide b exhibited an obvious cytotoxicity for cultured coeliac jejunal mucosa at a very low concentration (0.01 g/l) and was the most toxic peptide.

[Etiopathogenesis of celiac disease: the attraction and flimsiness of hypotheses]. / Etiopathogénie de la maladie coeliaque: attrait et fragilité des hypothèses.

The pathogenetic mechanisms which induce the intestinal lesions observed in coeliac disease are still unknown. The hypothesis of a primary intestinal peptidase deficiency has not been confirmed. The immunological theory is supported by a large number of findings, but it cannot explain all of the facts. The concept of a surface cell membrane abnormality aims at making up for the previous shortcomings, but it is based on very few unconfirmed data. Although the new pathogenetic theories for auto-immune diseases, and especially coeliac disease, are of considerable interest, experimental confirmation is eagerly awaited. Recently isolated pure toxic gliadin peptides may be of great value to test these new concepts in future research work.

Identification of a specific effector of the small GTP-binding protein Rap2.

Rap2 is a small GTP-binding protein that belongs to the Ras superfamily and whose function is still unknown. To elucidate Rap2 function, we searched for potential effectors by screening a mouse brain cDNA library in a yeast two-hybrid system using as a bait a Rap2A protein bearing a mutation of Gly to Val at position 12. This strategy lead to the identification of a protein that interacts specifically with Rap2A complexed with GTP, and requires an intact effector domain of Rap2A for interaction; we designated this protein Rap2-interacting protein 8 (RPIP8). Biochemical data obtained from in vitro studies with purified proteins confirmed the genetic results. Mouse RPIP8 consists of 446 amino acids, bears a coiled-coil domain between residues 265 and 313, and is expressed principally in brain. Its human counterpart, of 400 amino acids, exhibits 93.7% identity in their common region. A search for similar sequences in expressed-sequence-tags databanks revealed the presence in human and rodents of mRNAs encoding the 400-residue and 446-residue forms of RPIP8. Furthermore a doublet of 45-50 kDa, corresponding to the 400-residue and 446-residue forms of the protein, was detected by western blotting of mouse brain extracts and lysates from pheochromocytoma PC12 cells and the pancreatic beta-cell lines HIT-T15 and RIN-m5F. Using transient transfections of HIT-T15 cells it was possible to demonstrate that [Val12]Rap2 and wild-type Rap2 could be immunoprecipitated with RPIP8. These data therefore argue for RPIP8 being a specific effector of the Rap2 protein in cells exhibiting neuronal properties.

DNA polymorphism analysis of HLA class II genes in unrelated children and in first-degree relatives with type I diabetes.

Eighty unrelated diabetic children, seventy healthy controls and hundred and ten affected and unaffected first-degree relatives of twenty multiplex families were investigated by restriction fragment length polymorphism analysis of HLA class II genes using five probe/enzyme systems: DRB and DQB/Taq I, DRB and DQB/EcoRI and DQB/BamHI according to standard procedures described in the 10th Histocompatibility Workshop protocol. Comparison between the unrelated diabetic patients and the controls confirmed the positive association of type 1 diabetes with DR3(w17)DQw2 Dw24 or Dw25 and DR4DQw8 and the negative association with DR2(w15)DQw6, DR4DQw7 and DR7DQw2 haplotypes. In multiplex families, similar allele associations were found and the distinction between haplotypes present in diabetic patients and those that segregated to healthy family members allowed to observe striking differences between the "affected" and "unaffected" haplotypes, particularly for the subtypes of DR3(w17) DQw2, DR4DQw3 and DR2DQw1 haplotypes. Heterozygous siblings who carried both DR3DQw2 and DR4DQw8 subtypes disclosed a highly increased risk and more than 80% of DR3/DR4 affected siblings received a paternal DR4DQw8 together with a maternal DR3DQw2. These observations indicate that several genetic aspects influence susceptibility to type 1 diabetes: 1) some particular HLA class II subsets; 2) the parental origin of the predisposing genes; 3) the synergistic effect of both haplotypes, in particular DR3DQw2 and DR4DQw8. These results may help to better specify susceptibility markers for risk prediction in siblings.

Localization of the human oncogene SPI1 on chromosome 11, region p11.22.

Spi1 is an oncogene specifically activated in acute murine erythroleukemias induced by the Friend spleen focus forming virus (SFFV). Three probes were used for the chromosomal assignment of the human SPI1 oncogene: cDb1 and RaB2 correspond respectively to murine Spi1 and human SPI1 cDNA probes; C45a6B probe is a murine genomic DNA sequence located in the Spi1 5' region and is known as a major SFFV integration site in murine erythroleukemia cells. Somatic hybrid cells enabled cDb1 and RaB2 to be assigned to chromosome 11. The murine C45a6B probe, which is not included in the Spi1 gene, detected a homologous sequence on human chromosome 11. RaB2 was assigned to 11p11.22 by in situ hybridization. Three human genes known between 11p11 and 11p13 (FSHB, CAT, ACP2) were on murine chromosome 2. Therefore, the localization of human SPI1 on 11p11.22 was consistent with the assignment of the Spi1 oncogene to murine chromosome 2.

Identification and characterization of potential effector molecules of the Ras-related GTPase Rap2.

In search for effectors of the Ras-related GTPase Rap2, we used the yeast two-hybrid method and identified the C-terminal Ras/Rap interaction domain of the Ral exchange factors (RalGEFs) Ral GDP dissociation stimulator (RalGDS), RalGDS-like (RGL), and RalGDS-like factor (Rlf). These proteins, which also interact with activated Ras and Rap1, are effectors of Ras and mediate the activation of Ral in response to the activation of Ras. Here we show that the full-length RalGEFs interact with the GTP-bound form of Rap2 in the two-hybrid system as well as in vitro. When co-transfected in HeLa cells, an activated Rap2 mutant (Rap2Val-12) but not an inactive protein (Rap2Ala-35) co-immunoprecipitates with RalGDS and Rlf; moreover, Rap2-RalGEF complexes can be isolated from the particulate fraction of transfected cells and were localized by confocal microscopy to the resident compartment of Rap2, i.e. the endoplasmic reticulum. However, the overexpression of activated Rap2 neither leads to the activation of the Ral GTPase via RalGEFs nor inhibits Ras-dependent Ral activation in vivo. Several hypotheses that could explain these results, including compartmentalization of proteins involved in signal transduction, are discussed. Our results suggest that in cells, the interaction of Rap2 with RalGEFs might trigger other cellular responses than activation of the Ral GTPase.

Panel of twenty-five independent man-rodent hybrids for human genetic marker mapping.

To increase the efficiency of the human gene mapping, a panel of 25 man-rodent hybrids was selected out of 200 independent man-rodent hybrids produced in our laboratory since 1969. The hybrid panel was selected in such a fashion as to allow the asignment of a human marker M by respecting the two following complementary criteria: correlation between M and a chromosome and exclusion of the other chromosomes. The panel characteristics allowing the use of these two criteria were presented and discussed. A first series of enzyme markers were analysed to test the validity of the hybrid panel for the human gene mapping. The different possibilities and limits of the hybrid panel were also discussed, especially for the assignment of markers with DNA probes.

The human homologues of Fim1, Fim2/c-Fms, and Fim3, three retroviral integration regions involved in mouse myeloblastic leukemias, are respectively located on chromosomes 6p23, 5q33, and 3q27.

Three mouse genomic domains, Fim1, Fim2, and Fim3, were previously described as proviral integration regions frequently involved in the early stages of myeloblastic leukemogenesis induced in vivo or in vitro by the Friend murine leukemia virus. Fim2 was identified as the 5' end of the c-Fms protooncogene, which encodes the receptor of the macrophage colony stimulating factor (Csflr). The functions of Fim1 and Fim3 are not yet known, but these regions are highly conserved among different species. To examine whether these regions could correspond to known human loci involved in genetic alterations specific to some human leukemias, we undertook their chromosomal mapping. The localization of FIM2/c-FMS on 5q33 was confirmed. FIM1 and FIM3 were localized on human chromosomes 6p22.3-p23 and 3q27 respectively. Interestingly, translocations involving these two regions have been described in various hematopoietic malignancies: the t(6;9)(p23;q34) in acute nonlymphocytic leukemias and the 3q26-q28 translocations in a large variety of leukemias.

Localization of the active gene of aldolase on chromosome 16, and two aldolase A pseudogenes on chromosomes 3 and 10.

Southern blot analysis of human genomic DNA hybridized with a coding region aldolase A cDNA probe (600 bases) revealed four restriction fragments with EcoRI restriction enzyme: 7.8 kb, 13 kb, 17 kb and greater than 30 kb. By human-hamster hybrid analysis (Southern technique) the principal fragments, 7.8 kb, 13 kb, greater than 30 kb, were localized to chromosomes 10, 16 and 3 respectively. The 17-kb fragment was very weak in intensity; it co-segregated with the greater than 30-kb fragment and is probably localized on chromosome 3 with the greater than 30-kb fragment. Analysis of a second aldolase A labelled probe protected against S1 nuclease digestion by RNAs from different hybrid cells, indicated the presence of aldolase A mRNAs in hybrid cells containing only chromosome 16. Under the stringency conditions used, the EcoRI sequences detected by the coding region aldolase A cDNA probe did not correspond to aldolase B or C. The 7.8-kb and greater than 30-kb EcoRI sequences, localized respectively on chromosomes 10 and 3, correspond to aldolase A pseudogenes; the 13-kb EcoRI sequence localized on chromosome 16 corresponds to the aldolase active gene. The fact that the aldolase A gene and pseudogenes are located on three different chromosomes supports the hypothesis that the pseudogenes originated from aldolase A mRNAs, copied into DNA and integrated in unrelated chromosomal loci.

Assignment of the complement serine protease genes C1r and C1s to chromosome 12 region 12p13.

C1r and C1s are distinct, but structurally and functionally similar, serine protease zymogens responsible for the enzymatic activity of the first component of complement (C1). Recent comparisons indicate a significant degree of sequence similarity between C1r and C1s and support the hypothesis that they are related by gene duplication. Complementary DNA probes for human C1r and C1s do not cross-hybridize even at mild stringency conditions and are therefore gene-specific. Using a panel of 25 human-rodent cell hybrids, we have independently assigned the C1r and the C1s genes to chromosome 12. In situ hybridization analyses were consistent with these assignments, showing in addition that both C1r and C1s are located on the short arm of the chromosome in the region p13. These data suggest that the homologous C1r and C1s genes have remained closely linked after duplication of a common ancestor. The C1r and C1s loci also provide useful polymorphic DNA markers for the short arm of chromosome 12.

Localization of the acetylcholine receptor gamma subunit gene to human chromosome 2q32----qter.

The nicotinic acetylcholine receptor of skeletal muscle (CHRN in man, Acr in mouse) is a transmembrane protein composed of four different subunits (alpha, beta, gamma, and delta) assembled into the pentamer alpha 2 beta gamma delta. These subunits are encoded by separate genes which derive from a common ancestral gene by duplication. We have used a murine full-length 1,900-bp-long cDNA encoding the gamma subunit subcloned into M 13 (clone gamma 18) to prepare single-stranded probes for hybridization to EcoRI-digested DNA from a panel of human x rodent somatic cell hybrids. Using conditions of low stringency to favor cross-species hybridization, and prehybridization with rodent DNA to prevent rodent background, we detected a single major human band of 30-40 kb. The pattern of segregation of this 30-40 kb band correlated with the segregation of human chromosome 2 within the panel and the presence of a chromosomal translocation in the distal part of the long arm of this t(X;2)(p22;q32.1) chromosome allowing the localization of the gamma subunit gene (CHRNG) to 2q32----qter. The human genes encoding the gamma and delta subunits have been shown to be contained in an EcoRI restriction fragment of approximately 20 kb (Shibahara et al., 1985). Consequently, this study also maps the delta subunit gene (CHRND) to human chromosome 2q32.1----qter. In the mouse, the Acrd and Acrg genes have been shown to be linked to Idh-1, Mylf (IDH1 and MYL1 in humans, respectively) and to the gene encoding villin on chromosome 1. Interestingly, we have recently localized the human MYL1 gene to the same chromosomal fragment of human chromosome 2. These results clearly demonstrate a region of chromosomal homoeology between mouse chromosome 1 and human chromosome 2.

Chromosomal assignment of 14 genomic probes for highly polymorphic loci.

Over 500 probes revealing restriction fragment length polymorphisms (RFLPs) have been isolated by Schumm et al. (1988). We describe here the chromosomal assignment of 14 of the most highly polymorphic markers in that set of probes, with polymorphism information content values of up to 0.98. The probes were mapped using a panel of human x rodent somatic cell hybrids and were found to be distributed among nine different autosomes. Chromosome localization of such highly polymorphic markers has been an important step in the construction of the human genetic map, as a large number of RFLP probes has now been localized by genetic linkage studies to these loci.

The anonymous PAS45 probe detects RFLPs in 13q31.

An anonymous DNA probe PAS45 was isolated. This probe detects an RFLP with two alleles 1 and 2 at the same locus, with the different restriction enzymes (Bg1II, EcoRI, HindIII, PstI, MspI, XbaI). The observed polymorphism is explained by a chromosome rearrangement involving these enzyme cleavage sites. The frequency of alleles 1 and 2 was 0.875 and 0.125, respectively, in a sample of 48 unrelated individuals in France. Codominant inheritance of alleles 1 and 2 was demonstrated in 13 families with 30 offspring. The PAS45 probe was localized on chromosome 13 by somatic cell hybrid analysis and on 13q31 by in situ hybridization. The rearrangement on 13q31 is present in one out of four healthy individuals in France.

RGS14 is a novel Rap effector that preferentially regulates the GTPase activity of galphao.

In an attempt to elucidate the physiological function(s) of the Ras-related Rap proteins, we used the yeast two-hybrid system and isolated a cDNA encoding a protein that interacts with both Rap1 and Rap2, but not with Ras; the use of Rap2 mutants showed that this interaction is characteristic of a potential Rap effector. This protein was identified as RGS14, a member of the recently discovered family of RGS ('regulators of G-protein signalling') proteins that stimulate the GTPase activity of the GTP-binding alpha subunit of heterotrimeric G-proteins (Galpha). Deletion analysis, as well as in vitro binding experiments, revealed that RGS14 binds Rap proteins through a domain distinct from that carrying the RGS identity, and that this domain shares sequence identity with the Ras/Rap binding domain of B-Raf and Raf-1 kinases. RGS14 is distinguished from other RGS proteins by its marked preference for Galpha(o) over other Galpha subunits: RGS14 binds preferentially to Galpha(o) in isolated brain membranes, and also interacts preferentially with Galpha(o) (as compared with Galpha(i1)) to stimulate its GTPase activity. In adult mice, RGS14 expression is restricted to spleen and brain. In situ hybridization studies showed that it is highly expressed only in certain areas of mouse brain (such as the CA1 and CA2 regions of the hippocampus), and that this pattern closely resembles that of Rap2, but not Rap1, expression. Double in situ hybridization experiments revealed that certain cells in the hippocampus express both RGS14 and Galpha(o), as well as both RGS14 and Rap2, showing that the interaction of RGS14 with Galpha(o) and Rap2 is physiologically possible. Taken together, these results suggest that RGS14 could constitute a bridging molecule that allows cross-regulation of signalling pathways downstream from G-protein-coupled receptors involving heterotrimeric proteins of the G(i/o) family and those involving the Ras-related GTPase Rap2.

Assignment of the human fast skeletal muscle myosin alkali light chains gene (MLC1F/MLC3F) to 2q 32.1-2qter.

A DNA probe derived from a mouse intronless pseudogene including coding regions for the myosin fast skeletal muscle alkali light chains, MLC1F/MLC3F (suggested HGM symbol, MYL1), was tested on a panel of 25 independent man-rodent somatic cell hybrids in order to assign the human MLC1F/MLC3F gene to a human chromosome. A 3.7-kb TaqI human fragment was found to correlate with the presence of chromosome 2 in the hybrids, characterized both by cytogenetic analysis and reference enzyme markers. A regional assignment to 2q32.1-qter was possible using hybrids whose human parental strains bore a reciprocal translocation t(X;2) (p22;q32.1). The fact that IDH1 and the MLC1F/MLC3F gene are closely linked on chromosome 1 in the mouse and map to the same region of human chromosome 2 in man indicates, that these chromosomes have a conserved region of homology between them and that the human 3.7-kb TaqI fragment corresponds indeed to a functional gene.

Mapping of the gene for anti-müllerian hormone to the short arm of human chromosome 19.

The gene coding for human anti-Müllerian hormone (AMH) was localized to subbands p13.2----p13.3 on chromosome 19, using in situ hybridization and Southern blot analysis of a panel of man-mouse and man-hamster somatic cell hybrids.